LGMLJun 8, 2021

Adaptive Machine Unlearning

arXiv:2106.04378v1238 citations
Originality Highly original
AI Analysis

This addresses a critical security and privacy issue in machine unlearning for users who might delete data reactively, though it builds incrementally on prior non-adaptive work.

The paper tackles the problem of data deletion algorithms failing when deletion sequences are adaptive, i.e., chosen based on published models, by providing a general reduction using differential privacy to convert non-adaptive guarantees into adaptive ones, leading to flexible algorithms with strong provable guarantees.

Data deletion algorithms aim to remove the influence of deleted data points from trained models at a cheaper computational cost than fully retraining those models. However, for sequences of deletions, most prior work in the non-convex setting gives valid guarantees only for sequences that are chosen independently of the models that are published. If people choose to delete their data as a function of the published models (because they don't like what the models reveal about them, for example), then the update sequence is adaptive. In this paper, we give a general reduction from deletion guarantees against adaptive sequences to deletion guarantees against non-adaptive sequences, using differential privacy and its connection to max information. Combined with ideas from prior work which give guarantees for non-adaptive deletion sequences, this leads to extremely flexible algorithms able to handle arbitrary model classes and training methodologies, giving strong provable deletion guarantees for adaptive deletion sequences. We show in theory how prior work for non-convex models fails against adaptive deletion sequences, and use this intuition to design a practical attack against the SISA algorithm of Bourtoule et al. [2021] on CIFAR-10, MNIST, Fashion-MNIST.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes